Putting the F in FDA

Red no 3 is out

I’ve had a bunch of folks ask for me to comment on food dyes and their effects on health in the era of MAHA- they’ve been a hot topic for as long as I have been in the field of nutrition, with a lot of the buzz drum up by both early influencers (FoodBabes, etc) as well as the Center for Science in the Public Interest. The reasons for concern about food dyes are myriad, from believing corporations are putting unnecessary toxins in our food worsening kids behavior and causing cancer, to thinking that food dyes attract vulnerable populations (e.g. kids) to consume minimally nutritious ultraprocessed foods. My general take on them has been that the evidence is often mixed from human trials that they negatively effect readily measurable variables like behavior in kids, and that the Toxicology research rarely suggests that doses consumed by humans are at risk of being problematic (I’m much more concerned about the nutritional adequacy of the diet if someone is getting enough of a food dye in their diet to even approach being risky - think, eating bags and bags of M&Ms daily as a major source of calories). In the list of nutritional concerns for the American population, food dyes don’t typically come close to the top 10 — but they’re clearly an area where we see public interest and increasing bipartisan support for action (a rare thing with our dysfunctional Congress). That groundswell resulted in, this week, the FDA revoking the authorized use for one long-controversial food dye, Red No 3.

There’s not new data that popped up causing this revocation - the true rationale is almost certainly because of increasing public/political support for the FDA to actually focus on food (the FDA is regularly critiqued for having an underactive foods program, largely resulting from historical apathy and because it’s funding is quite limited). With the incoming potential of RFK Jr as head of HHS and the general enthusiasm from the MAHA movement causing a big interest in food/wellness on the right, typically perceived as a left crunchy granola Portalnd type of thing, there’s an obvious incentive for the FDA to do something that looks like it’s protecting our food from ‘toxic’ chemicals that companies put in it. (I would contrast this with the uptick in food safety issues, everything from E. coli and listeria outbreaks to the infant formula crisis and the DailyHarvest lentil crumbles debacle, that have generated limited if any public outcry or sweeping changes to food regs).

The History

So what’s the deal on Red No 3, apart from politics? It’s chemically known as erythrosine and is a xanthene dye that provides a red color. It was discovered by the Swiss chemist Karl Kussmaul at the University of Basel in 1876 and has been used as a dye broadly in the textile industry, for cosmetics and in food. There’s a long history to color additive regulation in the USA, dating back to the early 1900s when blatantly toxic compounds like lead & mercury were used as colorants. The Food and Drugs Act (1906), the establishment of the FDA in 1927, the Federal Food, Drug and Cosmetic Act (1938) and the Color Additive Amendments (1960) would continually update how color additives were regulated across industries. Red No 3 had been used since 1907 and received formal listing for uses in foods in 1969, following a petition by the Certified Color Manufacturer’s Association requesting permanent listing for use in food. In 1990, the FDA was petitioned to permanently authorize its use in cosmetics and topical drugs (it only had provisional authorization) but this petition was denied due to data in rats that it could cause cancer - the denial was based on the Delaney Clause, a provision that bans any carcinogenic additive regardless of safety tolerances (e.g. the dose of exposure doesn’t matter-it can just be banned because at some dose, it causes cancer). Shortly after in 1992, the FDA stated its intention to revoke the use of Red No 3 in food and drugs but the agency didn’t take action on this, citing limited resources required to remove the authorization. 30 years later, CSPI would submit a petition to the FDA and in November 2024, 23 bipartisan members of Congress would call on the FDA to remove its listing, resulting in the news this week.

On the surface, this whole history is perfect for fueling a narrative about the complete ineptitude of the FDA - a carcinogenic chemical was allowed to stay in the food supply for over 30 years with no action taken, despite alternatives being available. WTF?

It’s hard to argue against the perception that this is a flop from the FDA and there’s a clear need to refocus on the F in FDA. But resources are really at the heart of the issue here and we need an administration/Congress that will bolster the Foods Program (it’s overstretched trying to manage everything from food safety to health claims, enforcing supplement regulations, approving additives, etc). The narrative that the FDA was putting the American Public at risk, however, doesn’t hold a ton of weight (even though it definitely was useful for the political expediency of this revocation). The use of the Delaney Clause here is the tell here - it’s being revoked because of super high doses causing cancer in male rats (but not female rats or mice). There’s never really smoking gun data showing Red Dye No 3 is a carcinogenic risk - let’s dig into that.

The Toxicology

The data that caused the de-listing stir back in the 1990s were studies done in the 80s feeding various levels of Red No 3 to rodents. Prior to this time point, there were some toxicology studies undertaken but they used too few animals to be adequately powered and did not employ as rigorous histopathological methods to detect tumors - these led to the FDA in 1977 calling for new chronic toxicity studies. The Certified Color Manufacturers’ Association (CCMA) hired a contract organization to conduct chronic feeding studies (including in utero exposure) of Red No 3 in both male and female rats and mice. In CD-1 mice (60 male, 60 female per group) exposed to different levels (0, 0.3, 1 or 3%) of Red No 3 for 24 months, there was no evidence of cancer or other deleterious effects (reader note: these dye percentages are weight for weight, so 1% is 1g dye in 100g of food; these levels can also be expressed as dye intake per kilogram bodyweight as you’ll see below based on the concentration in the food * the grams of food eaten divided by the rodent weight in grams). In Sprague-Dawley Albino CD rats exposed in utero and up to 28 months to either 0, 0.1, 0.5, 1 or 4%, there were modest increases in thyroid weight for females at the 0.5 and 1% doses. For male rats, a higher incidence of combined thyroid follicular cell adenomas and carcinomas were observed at 0.1, 0.5 and 1 percent groups by the FDA (a finding disputed by CCMA). In the 4% group, a higher incidence of male rats with thyroid follicular cell adenomas (15/69 in the 4% vs 1/69 in the control) and with carcinomas (5/68 in the 4% vs 1/68 in the control) was found by the FDA (adenomas are considered more benign tumors potentially on the path to the more significant malignant tumor carcinoma phenotype). Another type of cancer, a C-cell adenoma was also higher in the 4% animals but there is substantial variability in their spontaneous occurrence in rats and was judged not to be biologically critical. Notably, at lower doses, there was no dose-response relationship between Red Dye Dose and adenoma or carcinoma number, leaving the 4% dose the real heavy hitter here showing evidence of oncogenicity and carcinogenicity. This lack of a dose-response relationship, some qualms about statistical analysis and the fact that the higher tumor burden in the 4% group did not result in increased mortality/differences in survival time were sticking points for the industry at the time disputing the FDA’s interpretation that Red Dye No 3 is carcinogenic.

The mechanism of the carcinogenicity was uncertain at the time - the dye does not accumulate in the thyroid and is not considered directly genotoxic; secondary mechanisms, such as hormonal disturbances, have long been considered and were heavily advocated by the industry at the time, who wanted to argue that the mechanism was due to high levels of the dye disrupting thyroid and pituitary hormones, effects that have a threshold effect and thus should not be regulated as a direct carcinogen. The FDA disputed the secondary mechanism data at the time of its review in the late 1980s but it is now well-accepted, with the World Health Organization’s Expert Committee on Food Additives (JECFA) ‘s most recent 2020 update of Red Dye No 3/Erythrosine reviewing the available data, reporting no concern for genotoxicity directly from the dye but consistent reports of it impacting thyroid hormone metabolism - namely, a reduced conversion of T4 → T3, causing an increased pituitary secretion of thyroid stimulating hormone (TSH), a hormone that is well-accepted in the rat to result in prolonged stimulus of the thyroid gland and cellular changes resulting in increased cancer risk. This effect occurrs rapidly at high doses (no effects at 0.03%, 0.06% or 0.5% but present at 4%, equivalent to ~17.5, 35.8, 298.3 and 4000 mg dye per kg body weight). To put these numbers into perspective for you, estimated intakes from both food surveys and total quantities of red dye certified to be used in the food supply estimate max intakes at 0.4 mg/kg bw, several orders of magnitude lower than the lowest doses causing no effect on thyroid hormone levels in humans (max exposures overstate things - typical intakes are estimated to be between 0-0.09mg/kg bw per day). The fact that effects in animals on thyroid hormones are observed so acutely has allowed for some experimental human evidence here, with studies showing very small changes in TSH levels after consuming 200mg per day for 2 weeks but no effect at 20 or 60 mg (all much higher than typical and extreme intake estimates). JECFA ultimately takes a conservative approach establishing an acceptable daily intake of 0-0.1mg/kg bw, ranges 60 fold lower than any effect observed in animals for any toxicological outcome.

A major consideration for organizations reviewing this thyroid cancer data is that the mechanisms of cancer in the rat don’t readily translate to humans - they clearly don’t even translate to mice or to female rats, which don’t show the red dye no 3-cancer link. Part of the problem for translating rats to humans is that the effect of red dye no 4 on reducing T4 → T3 and the resultant increase in TSH secretion by the pituitary has a lot of species specific context - namely, rats have a much shorter half life of T4 in their blood stream, due to differences in how tightly T4 binds to its binding protein. This results in increased T4 clearance rates in rats and much greater need to synthesize T4 on demand - to meet this increased demand, TSH levels 25X higher in rats than humans. Humans have a higher binding affinity of T4 for its globulin protein and thus a longer half life (you can think of this as having a better buffered, reserve of T4), and are much less sensitive to agents that reduce T4. This need to produce T4 by the rats results in many more cells in the thyroid that are metabolically active and are likely to take on a hyper proliferative state from excessive TSH stimulation, whereas human thyroid cells are much more quiescent. In sum, compounds that act on the pituitary-thyroid axis in the rat to increase thyroid cancer risk likely don’t translate to humans all that well, if at all.

When considering the known species differences in thyroid function and the very high doses of Red Dye No 3 needed to elicit impacts on thyroid function and produce thyroid adenomas and carcinomas in only male rats, you can start to understand why a FDA with limited resources did not necessarily rush to remove what is being painted as a ‘threat to public health’ in the current media cycle. Again, the risk to health would be from the very poor quality diet likely being consumed from high red dye no 3 intakes, not risky intake levels of the dye itself.

That all said, I’m ultimately hopeful that this bit of action by the FDA is a sign of change that it’s going to more broadly take on challenges facing us in the food system and that its not simply performative for a hostile administration - time will tell. My hope is that the action will be on cross cutting issues related to the food supply that represents a real risk to acute and chronic health, not just a single food dye at a time.

For those in this for the regulation of chemical additives side, I think it's worth being critical of this ‘win’ - there are thousands of additives in the food supply. They're all not going to go down via a Delaney Clause. Going 1 by 1 would take an insane amount of effort on the part of advocates and an underfunded FDA. Change on food additives needs to come at the level of how things are approved that ensures a safer food supply across the board, across all (or close to all) additives. I'm hoping to see more productive conversations from experts and advocates on BigThink areas for regulatory change.